Otras propiedades

activity in the potentiation of ciprofloxacin (4-fold). The introduction of substituents in the phenyl ring of compound

90 led to unsatisfactory results, with the best compound

displaying moderate antibacterial activity. A methyl group in the C-5 position did not show influence on the activity, and the modification of the ether linkage led to varied effects: introduction of a sulfanyl, thus conserving the H-bonding acceptor capacity, led to similar potentiating activity, whereas the introduction of an amine linkage, an H-bond donor, led to a decrease in the activity. Finally, the shift of the benzyloxy group from the C-6 to the C-5 resulted in no activity [110].

Compounds AE-848/42434549 (94) and AN-465/42885978 (95) were hits in a virtual screening of novel NorA efflux pump inhibitors in S. aureus. The structures deemed as active in silico were later synthesized and evaluated on their ability to prevent ethidium bromide efflux, as well as ciprofloxacin potentiation. Compound 94 and 95 were as potent as reserpine (30) for ethidium bromide efflux and more potent at synergizing with ciprofloxacin. These hit compounds 94 and 95 paved way for the synthesis of new analogues, which allowed SAR studies to be made, in order to assess which parts of their structure were determinant for the inhibition of NorA. For compound 94, it was shown that limited modifications to the sulfone and amide group do not cause significant changes in activity. On the other hand, modifications to the imidazole ring result in a reduction of activity. Replacement of the 1-methylpyrrole ring with a furan also shows a decrease in the activity of the compound. As for compound 95, the O-benzyl group is essential and the introduction of a benzyl in the secondary nitrogen leads to a complete loss of activity. Furthermore, the methyl group of the 2-(1-phenyl-1-propanol) moiety can be replaced with a 2- (1-phenyl-1-ethanol) portion without a decrease in the activity [111].

3.1.3. Existing Drugs

Several drugs already available for the treatment of various diseases have shown potential to be used as MFS EPIs. Reserpine (30), mentioned before, is one of them. Other drugs include verapamil, omeprazole, paroxetine and chlorpromazine. Even though bacterial MFS transporters and mammalian ABC efflux systems, as P-gp, lack structural homology, they share similar substrate profiles, which means that mammalian MDR inhibitors can also present bacterial efflux inhibition. However, the concentrations needed for efflux pump inhibition are too high, which means that toxic effects can arise if these compounds are used for this purpose [18], [112]. In fact, the addition of verapamil to chemotherapy was accompanied with cardiac arrhythmia and hypotension [113]. This leads to the requirement of developing analogues of these existing drugs maintaining the desired EPI activity at concentrations that do not display toxicity [112].

The antiarrhythmic verapamil, a known P-gp inhibitor, has shown moderate NorA inhibitory activity in both B. subtilis

and S. aureus, enhancing fluoroquinolone activity [114-115]. However, the best results were obtained with proton pump inhibitors omeprazole (96) and lansoprazole. Therefore, a series of pyrrolo[1,2-a]quinoxaline derivatives mimicking the omeprazole structure (96) were designed, with the aim of evaluating their EPI activity against NorA. All the eleven derivatives showed a reduction in norfloxacin’s MIC, being compound 97 the most active, reducing the MIC by 16-fold at 128 µg/ml. These compounds did not show disturbance of the electrical potential and the transmembrane pH, excluding the alteration of the proton-dependent pump as the mechanism of action. Instead, it is thought that these compounds interact directly with NorA. It was shown that the benzimidazole moiety is critical for the activity of these compounds (Figure 10), and replacing it with an imidazole, a pyridine or a pyrrolo[1,2-a]quinoxaline ring resulted in loss of potency. Furthermore, if there is a methoxy- substituent in the benzimidazole ring, potency increases. Introduction of chlorine in the C-7 of the pyrroloquinoxaline heterocycle provided an efficient restoration of norfloxacin’s bactericidal activity. Finally, the benzimidazolyl group was also important for activity [112].

This was not the first time that pyrrolo[1,2-a]quinoxaline derivatives were described as EPIs. In fact, a previous study by the same group produced 12 new 4-[116]pyrrolo[1,2-

a]quinoxalines and one pyrrolo[1,2-α]thieno[3,2-e]pyrazine

compound. Two compounds, 98 and 99, were more active than reserpine (30), diminishing norfloxacin’s MIC by 16- fold in a NorA overexpressing strain of S. aureus. This study showed that 4-[116]pyrrolo[1,2-a]quinoxalines and their 7- methoxy analogues were more active, in contrast with the chlorine- and 8-phenyl-substituted compounds, that presented the lowest effects. Replacement of the N,N- diethylamino group by pyrrolidine, its isomer with restricted conformation, led to enhanced EPI activity, whereas replacing it with a piperidine ring led to more flexible and, consequently, less active compounds. The pyrrolothienopyrazine 99 moiety yielded the best results, suggesting that sulfur, an electron-rich atom, potentiates the EPI activity [116].

GG918 (19) was discovered in a screening program aimed to identify inhibitors of mammalian P-gp, and was able to increase the concentrations of paclitaxel. This compound did not present antibacterial activity against the strains of S.

aureus tested. However, the co-administration with

norfloxacin resulted in a 4-fold reduction in the MIC of the NorA overexpressing strain [52].

Two mammalian MDR inhibitors, biricodar (100) and timcodar (101), confer increased drug sensitivity to cells expressing both the P-gp and multidrug resistance protein (MRP)-1 efflux systems. They also show potentiation of multiple antibiotics in Gram-positive bacteria. In fact, both compounds were able to potentiate the activity of ethidium bromide against S. aureus, E. faecalis, and S. pneumoniae. For the specific case of S. aureus, both compounds were able to partially reverse fluoroquinolone resistance. The fact that the MIC of ethidium bromide was non-saturable for the three

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59

For Review Only

Short Running Title of the Article Journal Name, 2014, Vol. 0, No. 0 11

bacteria tested suggests that 100 targets multiple efflux pumps [117].

A series of polysubstituted pyrroles was screened for their P- gp inhibition activity, which made way for the identification of a lead, and the compounds were further investigated for their bacterial efflux pump inhibition activity. A simplified pyrrole alkaloid, 102, displayed dual activity as both a P-gp and a NorA inhibitor against S. aureus. This compound led to 4-fold improvement in the MIC of ethidium bromide in the NorA overexpressing strain. Compound 102 was also able to reduce ciprofloxacin’s MIC by 8-fold when at 50 µM [118].

Tariquidar (103), a third generation MDR modulator, with activity in inhibiting P-gp and BCRP efflux systems, has shown promise in inhibiting efflux pumps in S. aureus. In fact, this compound showed an increase of the isotopic form of ciprofloxacin in all investigated strains, displaying best results in the strain that overexpresses the NorA pump. [119] The class of phenylpiperidine selective serotonin reuptake inhibitors (PSSRI), which includes the antidepressant paroxetine (104), showed efflux-related potential against S.

aureus, particularly that conferred by NorA. In fact,

paroxetine and its isomer, 105, showed potency with respect to inhibition of this efflux pump against ethidium bromide. The isomer 105 also interfered with the accumulation of norfloxacin by NorA, reducing by 4- to 8-fold its MIC. [120] Femoxetine (106) is a paroxetine derivative, with activity as an EPI. [120] Studies were performed as to determine which moieties in paroxetine- and femoxetine-like PSSRI were detrimental for efflux pump inhibition in S. aureus, and conclusions from SAR are highlighted in Figure 11. Synthesis of derivatives showed that the fluorine atom on the phenylpiperidine moiety is not associated with the activity. The N-substitution of the piperidine ring yields different kinds of activity, suggesting that the NH group forms a positive binding contact with the target pump, which does not happen if there is a N-substitution (compound 106). The phenyl ether substituents also play an important role as determining the EPI activity. [121]

Also for this kind of compounds, a study was performed, in order to conclude whether the 4-phenyl ring is required for EPI activity and the effects of changes in the two-atom aryloxymethyl linker, replacing it with a thioether, amine, and alkene linker portions. Compounds 107 and 108 displayed the lowest MIC against S. aureus. Synthesis of multiple derivatives proved that the 4-phenyl moiety is not vital for EPI activity and that a variety of two-atom linker groups for the 3-aryl piperidine moiety can maintain EPI activity [122].

The ciclooxigenase-2 (COX-2) inhibitor celecoxib (109) has also received attention for its MDR inhibitory activity. There had been a link between COX-2 and mdr1 gene expression, which has consequences in cancer chemo resistance, and the use of celecoxib (109) reversed drug resistance. If this is true for mammalian efflux pumps, for bacteria there is no evidence of a COX-2-like gene, suggesting that celecoxib acts differently in inhibiting MDR in bacteria. A study was carried out in order to prove this hypothesis, which

demonstrated that this drug helped in increasing sensitivity to antibiotics in S. aureus and Mycobacterium smegmatis. These effects probably occur due to the blockage of MDR transporters involved in the efflux of antibiotics, such as ampicillin, ciprofloxacin, kanamycin, and chloramphenicol. The observation of the accumulation of ethidium bromide intracellularly leads to the conclusion that this compound acts as a NorA inhibitor [123].

As celecoxib (109) demonstrated such promising activity, analogues were tested, which bore low or none anti- inflammatory activity and possessed the 1,4- dihydropyrazolo[4,3-c]-benzothiazine-5,5-dioxide nucleus. Ethidium bromide efflux inhibition was tested first in silico, and the 17 molecules that fit the chemical space requirements were then reduced to six, based on the activity prediction. These compounds were then screened for their in

vitro NorA inhibitory activity, using an overexpressing strain

of S. aureus. One derivative in particular, compound 110, showed the best activity concerning ethidium bromide efflux (76.9%) which is better than celecoxib (109) (64.5%). When in combination with ciprofloxacin, 110 displayed results comparable to reserpine (30) and better than paroxetine (104), which makes it a lead compound for the development of inhibitors of this kind [124].

Phenothiazine and thioxanthene derivatives are currently used as neuroleptic and antiemetic agents, and have been described as modest, but broad, antibacterial agents. Although their antibacterial activity is not enough to be clinically relevant, they have shown synergy with standard antibiotic agents, and have also been associated with the inhibition of eukaryotic MDR efflux pumps, such as P-gp. The mechanism through which these compounds exert their antimicrobial potentiation is not yet fully understood. However, it is thought that it might have to do with the inhibition of efflux pumps [125]. A study focusing on the activity of several compounds of these families against S.

aureus was carried out, with the main aim of showing their

potential in inhibiting NorA. The compounds tested, chlorpromazine, fluphenazine, thioridazine, prochlorperazine (111), cis(Z)-flupentixol, and trans(E)-flupentixol (112), displayed intrinsic antimicrobial activity and good results when ethidium bromide efflux was assayed against a NorA overexpressing strains of S. aureus. Additionally, phenothiazine 111 and thioxanthene 112 were also able to reduce the proton motor force of S. aureus, by reducing the transmembrane potential [125].

Since the phenothiazine moiety proved itself to work as a template for the synthesis of new MDR EPIs, it was chosen to be the scaffold for derivatives with improved activity. The logic for these studies lied in the elimination of the structural features that were responsible for neuroleptic activity. As such, drastic modifications were made, such as the elimination of one ring of the tricyclic benzothiazine backbone and of the chain linked to the N-10 atom, a tertiary amine whose protonation is crucial for the interaction with the dopaminergic receptor, and the addition of a substituted phenyl ring at the C-3 position, to guarantee better lipophilicity. Therefore, a prototype of new 3-phenyl-2H-

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59

For Review Only